ARIMAX Modeling of Hive Weight Dynamics Using Meteorological Factors During Robinia pseudoacacia Blooming

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

Recommendations for Improvement：

1. Further solve the nonlinear relationship between the change in beehive weight and temperature. This study points out that there is a nonlinear relationship between the change in beehive weight and the temperature. The current ARIMAX model is a linear model and may not be able to fully capture this complex nonlinear dynamic.
   Modification suggestion: Explore and apply nonlinear models or hybrid models, such as deep learning models (such as CNN, ANN, LSTM) or other time series models (such as VAR, VEC), to improve prediction accuracy, especially during the peak period of collection activities.
2. Incorporate more underrepresented biological and environmental factors. The paper mentioned that underrepresented biological and environmental factors lead to large fluctuations in collection activities, which is a limitation of the current method.
   Modification suggestion: If possible, include more detailed internal data of bee colonies (such as bee colony temperature, humidity, sound, CO2, bee colony flow, etc.) or other more microscopic environmental data, as well as more detailed plant phenology information, to more comprehensively reflect the complex interaction between bee colony activities and the environment.
3. Improve the ability to handle beekeeping interventions and abnormal events. Although the study selected undisturbed collection period data, in actual applications, beekeeping interventions (such as honey collection, feeding, treatment) and extreme weather events occur frequently, which can lead to abnormal changes in beehive weight.
   Suggestions for improvement: Integrate more robust data preprocessing techniques such as anomaly detection or breakpoint analysis to automatically identify and process weight changes caused by these non-collection activities, making the model more robust in a wider range of practical application scenarios.
4. Consider more explicit seasonal component modeling. The paper mentioned that the beehive weight has a strong daily cycle pattern, and the average daily cycle pattern is used as a regression factor in the model. At the same time, the paper also suggests adding a seasonal component to this strong daily cycle behavior.
   Suggestions for improvement: According to our research results, the changes in bee colony weight and climate change show obvious seasonal characteristics. This strong daily cycle (regarded as a short-term seasonality) can be more explicitly modeled in other time series models to see if the model's fitting and prediction capabilities can be further improved.

 

Writing and Presentation Suggestions：

1. Suggestions for writing/content: Although the overall language is clear, some sentences could be more refined and less repetitive. For example, when describing the purpose, methods, or conclusions of the study, make sure that the word choice best reflects the content of the study.
2. Suggestions for improvement in the method section: In the method section, a more detailed explanation of why the ARIMAX (12,0,6) model was chosen (line 256) could be provided. A deeper explanation of the specific meanings of p=12, d=0, q=6 (e.g., 12-hour autoregression, no difference, 6-hour moving average) and how they reflect the characteristics of the changes in beehive weight would help readers understand the rationale for the model selection. For example, p=12 may be related to some biological or environmental cycle, or reflect the cumulative effects of the past 12 hours.
3. Further explain the specific way and significance of incorporating the average circadian rhythm pattern as a regressor into the model. This section is one of the key innovations of the model, and it could be emphasized more on how this effectively captures the daily fluctuations in beehive weight.
4. In the discussion, when temperature is mentioned as the most important predictor and is significant in most bee colonies, the biological mechanisms behind this can be explored in more depth (such as the temperature threshold and optimal temperature range for bees to leave the nest to collect honey and its direct guiding significance for beekeeping practices).
5. In the discussion section, consider whether strong daily cycles or seasonal components can be modeled more explicitly in time series models, rather than just incorporating them as regressors into the average pattern.

Author Response

Please see the attachment.

 

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

In the introduction, we recommend to the authors divide and organize the paragraphs, as they are long and very tiring for readers. We also suggest to the authors that introduction have five to eight paragraphs.

In the discussion, we recommend to the authors discuss better the data presented with the literature, avoiding adding or comparing data not presented in the work and correlating them with the literature.

The conclusion must be clear and objective, and correlated with the objective of the research. We recommend to the authors to check the second paragraph of the conclusion, as it should be removed because it deals with materials, methods and results, and not the conclusion.

Author Response

In the introduction, we recommend to the authors divide and organize the paragraphs, as they are long and very tiring for readers. We also suggest to the authors that introduction have five to eight paragraphs.

In the discussion, we recommend to the authors discuss better the data presented with the literature, avoiding adding or comparing data not presented in the work and correlating them with the literature.

The conclusion must be clear and objective, and correlated with the objective of the research. We recommend to the authors to check the second paragraph of the conclusion, as it should be removed because it deals with materials, methods and results, and not the conclusion.

ANSWER: Thank you for these valuable recommendations. We have revised the manuscript accordingly. The introduction has been restructured into five well-organized paragraphs to enhance readability and clarity. The discussion section was refined to better align our findings with relevant literature, also incorporating suggestions from other Reviewers. Additionally, the conclusion was revised to be more concise and directly connected to the study’s objectives and main findings.

Reviewer 3 Report

Comments and Suggestions for Authors

In this study, the authors measured the changes in beehive weight during the honey flow period of Robinia pseudoacacia (black locust) from April 15 to June 15 using sensors. They used an ARIMAX model to assess the relationship between some meteorological factors in Hungary, such as 2-meter actual temperature (t), 1-hour minimum temperature (tn), 1-hour maximum temperature (tx), surface temperature (tsn), radiation (sr), wind speed (fs), wind gust (fx), precipitation (r), and relative humidity (u), and the changes in beehive weight during the Robinia pseudoacacia honey flow period. This study provides insights for beekeepers regarding the impact of weather on the honey flow of Robinia pseudoacacia and the influx of nectar into the beehives, enabling them to develop beekeeping management strategies based on weather variations. Therefore, this article holds significant value for publication in a scientific journal. However, there are some flaws in the manuscript that need to be revised to enhance the quality of the article and meet the standards for publication. Here, I would like to pose a few academic questions to the authors regarding this paper, and I hope for an engaging discussion.

1. What is an ARIMAX model?
2. This article describes the value of the technology and theory, how valuable is the practical application? Is there hope for future use in beekeeping? Is the economic cost affordable to the beekeeper?
3. How strong is the colony, and how many combs are in the hive?
4. Were all bee colonies subjected to the same beekeeping practices during the experiment?
5. The authors were asked to explain why the ARIMAX model can be used to assess the effect of meteorological factors on hive weight changes during heavy nectar flows.

 

If the author addresses all the issues raised in the manuscript, then this paper will be worth publishing in atmosphere. I recommend the authors make minor revisions.

 

Title

This section is fine.

 

Abstract

Line 17: “0.1and 0.3 kg/hour” A space is missing between “0.1” and “and”.

 

Introduction

Lines 25: “Apis mellifera L., commonly known as a honey bee”

To the best of my knowledge, Apis mellifera L., commonly known as the Western honey bee.

 

Materials and Methods

The Materials and Methods section is deficient in standardized information pertaining to the bee colonies, including details such as colony size (colony strength) and health.

 

Results and Discussion

From what I understand, these sections are fine.

 

Conclusions

 

Line 364-386: The conclusion is too lengthy and overly verbose. It needs to be summarized concisely for clarity.

 

References

Italicize genus names:

line 446: “Robinia Pseudoacacia”

line 449: “Apis Mellifera”

Please check the entire reference list for proper formatting.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 4 Report

Comments and Suggestions for Authors

The article “ARIMAX modeling of hive weight dynamics using meteorological factors during Robinia pseudoacacia blooming” provides data to estimate hourly hive weight variation by applying linear time-series models to hive weight data collected from active apiaries during intensive foraging periods, considering atmospheric predictors. The peculiarity of the article is the use of ARIMAX modeling. The novelty of the study raises some questions, since many studies have previously investigated the influence of various weather factors. Also, the predominant influence of temperature on the accumulation of hive weight was previously determined.
The article can be published in the journal “Atmosphere” after minor revisions.
Line 75. The frequency of weighing should be described in more detail. Since weight is a very important parameter in this article. If they weighed every 10 minutes, then the mention of 30 and 60 minutes is unclear.
The authors write that nectar was mainly collected from Robinia pseudoacacia (black locust). The discussion of the choice of this particular plant and the conclusions about the role of the plant are not given in the text of the article.
The year of manufacture, the name of the company and the country should be given for each device used.
Please tell me, what was the distance between the different apiaries? I ask for a more detailed explanation of the location of the weather station and the apiaries relative to each other. How many weather stations transmitted weather data?
Table 1. Please provide a row with data for 1991-2020 that was used for comparison.
Formula 1 and lines 143-147. Please provide a decoding of the symbols used in the order they are mentioned.
Figure 1. Using different colors, rather than 4 shades of green, will make the drawing clearer.
Page 160. Please clarify point 4)
Page 187. Please clarify the situation, why the error is greater than the basic value (1.80 ± 1.85)
Table 2. How can the parameters for hives 12, 17, 18 be explained?

Author Response

Reviewer #4:

Comments and Suggestions for Authors

The article “ARIMAX modeling of hive weight dynamics using meteorological factors during Robinia pseudoacacia blooming” provides data to estimate hourly hive weight variation by applying linear time-series models to hive weight data collected from active apiaries during intensive foraging periods, considering atmospheric predictors. The peculiarity of the article is the use of ARIMAX modeling. The novelty of the study raises some questions, since many studies have previously investigated the influence of various weather factors. Also, the predominant influence of temperature on the accumulation of hive weight was previously determined.

The article can be published in the journal “Atmosphere” after minor revisions.

ANSWER: Thank you for your thorough review and the positive evaluation of our manuscript. 

Line 75. The frequency of weighing should be described in more detail. Since weight is a very important parameter in this article. If they weighed every 10 minutes, then the mention of 30 and 60 minutes is unclear.

ANSWER: Thank you for your observation regarding the frequency of hive weighing. We modified this line in the Methodology section for better understanding.

The authors write that nectar was mainly collected from Robinia pseudoacacia (black locust). The discussion of the choice of this particular plant and the conclusions about the role of the plant are not given in the text of the article.

ANSWER: Thank you for pointing out the lack of explanation regarding the choice of this specific plant. In response, we have added a brief discussion in the Discussion and Conclusion section in the revised manuscript to clarify this point. Robinia is the most important bee pasture in our region with large agricultural and economic importance. This focus provided widely available hive weight data from beekeepers, and also enhances the practical applicability of our development. Furthermore, its relatively short blooming period and immense nectar flow clearly defines the foraging period without overlapping with other pastures, which makes it a good subject for time series modeling purposes.

The year of manufacture, the name of the company and the country should be given for each device used.

ANSWER: Thank you for your comment. We fully understand the importance of transparency in scientific reporting. However, in this case, the manufacturer has requested not revealing the company name or specific identifying details of the apiaries to protect the identity of collaborating beekeepers. We confirm that the hive scale manufacturer and all the apiaries involved in this study are located in Hungary. We kindly ask for your understanding regarding this matter.

Please tell me, what was the distance between the different apiaries? I ask for a more detailed explanation of the location of the weather station and the apiaries relative to each other. How many weather stations transmitted weather data?

ANSWER: Data from six different weather stations were used to characterize the environmental conditions relevant to each apiary. The mean distance between the apiaries and their associated weather stations was 10 km, with the minimum distance being 1.4 km and the maximum distance 18 km. Regarding the spatial distribution of the apiaries, the minimum distance between any two apiaries was 3 km, the median distance was 160 km, and the maximum distance was 424 km. Although we cannot reveal the exact location of the apiaries due to data restrictions, we have now included a more detailed description of the spatial setup in the revised manuscript.

Table 1. Please provide a row with data for 1991-2020 that was used for comparison.

ANSWER: Thank you for your suggestion. We have added a row to Table 1 containing the data for the 1991–2020 reference period in Hungary used for comparison, as requested.

Formula 1 and lines 143-147. Please provide a decoding of the symbols used in the order they are mentioned.

ANSWER: Thank you for your helpful suggestion. In response, we have revised the text to include a detailed decoding of the symbols used in Equation (1), in the order they appear.

Figure 1. Using different colors, rather than 4 shades of green, will make the drawing clearer.

ANSWER: Thank you for your comment regarding the color scheme in Figure 1. We understand the concern about visual clarity. However, we intentionally used various shades of green to represent the individual years in a slight way, as our focus is on the red line, which indicates the average. The green lines are meant to demonstrate the interannual variability without focusing on any invidividual year and with no distraction of the mean (red) line.

Page 160. Please clarify point 4)

ANSWER: Thank you for pointing this out. We have revised and clarified point 4) on line 160.

Page 187. Please clarify the situation, why the error is greater than the basic value (1.80 ± 1.85)

ANSWER: Thank you for your observation. We agree that the reported standard deviation being greater than the mean value may raise questions.  Upon reviewing the data, we identified a minor error related to data preprocessing, which has since been corrected. We re-ran all model calculations and statistical summaries to ensure accuracy. It is important to note that, hive weight can fluctuate significantly depending on weather conditions, nectar availability and colony strength. As a result, some days exhibit minimal gain or even loss, while others show large increases, such as the observed maximum of 9.27 kg/day.

Table 2. How can the parameters for hives 12, 17, 18 be explained?

ANSWER: Thank you for your insightful question regarding the model parameters for Hives 12, 17, and 18. As previously mentioned, upon reviewing the data, we identified and corrected a minor issue affecting our calculations. After addressing this, we reprocessed the relevant datasets and updated the model outputs for all hives, including Hives 12, 17, and 18. The revised results are now reflected in the updated manuscript, and we have ensured consistency across all figures and tables.

Round 2

Reviewer 2 Report

Comments and Suggestions for Authors

this manuscript got much better after the fixes and doesn't even look the same;

the introduction has a very good theoretical fundamentation; the material and methods are very detailed and the statistical analysis is correct too;

the conclusions are according the subjects;

Author Response

Thank you for your time dedicated for this review and the positive evaluation of our manuscript.